1. Field of the Invention
The present invention relates to an image processing apparatus, an image forming apparatus, an image processing method and a computer-readable recording medium that determine whether a document to be read is a double-sided document or a single-sided document.
2. Description of the Related Art
In a copier or image reading apparatus, when a double-sided document is copied by an automatic document feeding apparatus, etc., or a document image is sent by electronic mail to a predetermined e-mail address as scan to e-mail, if a plurality of documents to be read include both single-sided documents and double-sided documents, first, a user needs to separate the documents into single-sided documents and double-sided documents by hand and then needs to perform reading operation on respective documents. Hence, the user needs to rearrange outputted documents in the original order.
On the other hand, if reading is performed without thus separating the documents, unwanted blank sheets are outputted. Particularly when color blank sheets are mixed, color toner (ink) is to be wastefully consumed. Also, when read document image data is stored, storage space of a hard disk or the like is to be wastefully used.
To solve the problems described above, the following methods are proposed.
In an invention described in Japanese Unexamined Patent Publication JP-A 6-245003 (1994), in a file of documents including both double-sided documents and single-sided documents, distinction is made between a double-sided document and a single-sided document by the absence or presence of image. Determination on whether an image is absent or present is made such that the number of black pixels is counted to check whether a count value exceeds a certain reference value and if exceeds, then it is determined that an image is present. Alternatively, such determination is made based on the amount of compressed data of image.
In an invention described in Japanese Unexamined Patent Publication JP-A 9-261383 (1997), when making a distinction between a double-sided document and a single-sided document, a threshold is set to a level close to white and the number of pixels that are less than the threshold is counted, whereby determination is made on presence or absence of an image. When the above-described counted number is greater than a predetermined value which is set in advance, it is determined that an image is present.
In an invention described in Japanese unexamined Patent Publication JP-A 9-284490 (1997), a binary level of data of image is found and the number of pixels of the binary level is compared with a predetermined value to determine whether the data of image corresponds to a blank page document.
When the above-described conventional techniques are used, isolated dots which are not present in data of image, the influence of noise upon reading, and moreover, stains on a page background portion and dust attached to the page background portion are counted as black pixels and thus the accuracy of blank page image determination decreases. This occurs in both of the case of counting white pixels and the case of counting black pixels.
Furthermore, since the conventional techniques do not consider a color blank page document at all, determination cannot be made on a color blank page image.